Hi!
I am looking for a good complementary pair of transistors for my VAS stage (rail voltage is +/-42V). I have searched the forum and it seems that the MJE340/350 are mostly used, but considered not to be the best... I have found some other BF/Sanyo/etc. parts, but most of them seem to be discontinued or I don't find a reseller... further, I would be very happy to get the Spice models for the transistors somewhere...
Any suggestions?
Thanks,
stegmaie
I am looking for a good complementary pair of transistors for my VAS stage (rail voltage is +/-42V). I have searched the forum and it seems that the MJE340/350 are mostly used, but considered not to be the best... I have found some other BF/Sanyo/etc. parts, but most of them seem to be discontinued or I don't find a reseller... further, I would be very happy to get the Spice models for the transistors somewhere...
Any suggestions?
Thanks,
stegmaie
Hi,
Also depends on the VAS topology, eh?
For a start, BD139/BD140 aren't that bad and cost essentially nothing.
Judging from datasheets and models KSA1381/KSC3503 look fine. But unless odering a larger quantity from a Fairchild distributor, you would most likely only get 2SA1381/2SC3503 and then you must hope, that the originals are as good as the copies.
http://www.fairchildsemi.com/models/PSPICE/Discrete/Bipolar_Transistor.html
Panasonic 2SA794/2SC1567 is available through Reichelt and not bad at all.
10mA Iq enough? Panasonic 2SA1124/2SC2632 in 'stretched' TO-92 will do. Several german mail-order stores have them.
Moderately nostalgic feeling: BC141/BC161 in TO-39 metal case!
Prefer more power dissipated: Sanyo 2SA1011/2SC2344 or Sanken 2SA1488/2SC3851 in TO-220AB/TO-220FM
Regards,
Peter Jacobi
Also depends on the VAS topology, eh?
For a start, BD139/BD140 aren't that bad and cost essentially nothing.
Judging from datasheets and models KSA1381/KSC3503 look fine. But unless odering a larger quantity from a Fairchild distributor, you would most likely only get 2SA1381/2SC3503 and then you must hope, that the originals are as good as the copies.
http://www.fairchildsemi.com/models/PSPICE/Discrete/Bipolar_Transistor.html
Panasonic 2SA794/2SC1567 is available through Reichelt and not bad at all.
10mA Iq enough? Panasonic 2SA1124/2SC2632 in 'stretched' TO-92 will do. Several german mail-order stores have them.
Moderately nostalgic feeling: BC141/BC161 in TO-39 metal case!
Prefer more power dissipated: Sanyo 2SA1011/2SC2344 or Sanken 2SA1488/2SC3851 in TO-220AB/TO-220FM
Regards,
Peter Jacobi
Stegmaie,
As general rule of thumb, the higher the VCE the transistors can handle and the higher power they can dissipate, the lower their gain and the lower their GBW product. So, from a performance point of view for +/- 42 V, you should take the one that has a Vce limit of say 100 or 120V, but no more, and that can easily handle the power required, but not very much more.
Jan Didden
As general rule of thumb, the higher the VCE the transistors can handle and the higher power they can dissipate, the lower their gain and the lower their GBW product. So, from a performance point of view for +/- 42 V, you should take the one that has a Vce limit of say 100 or 120V, but no more, and that can easily handle the power required, but not very much more.
Jan Didden
stegmaie said:
Hello everybody, this is my first post.
I would recommend using bf871/872 pair. They can withstand very high voltages, colector base capacitance is very low (2 or 3 pf), and ft is 60 Mhz. Packaging case is TO 202, which means that they can dissipate 300 or 400mW without overheating. Don't use TO 92 and TO 39 devices, they would probably overheat (unless your VAS current is few milliamps). BF471/472 offers similar caracteristic in TO 126 package.
Dule
Stegmaie,
Speed and low base/collector capacitance is important.
As Jan says, fix the voltage a little over twice the rail voltage; for a 42V rail 100-120V is fine for Vceo.
The higher the beta, the more the open loop gain of the amp, and the more is the feedback factor. Generally, this is good. So try to find a device no slower than 100MHz and a beta of 60 minimum.
I have used 2SC1819 and 2SC3955 with good results, but both these devices are now obsolete. Sanyo have a very good manufacturing process which gives high speed, low Cbc, and good beta. IMHO the voltage amp is the pivotal stage in a power amplifier. Watch the dissipation; no more than one third the rating for the device, and keep temperatures low for reliability.
Cheers,
Hugh
Speed and low base/collector capacitance is important.
As Jan says, fix the voltage a little over twice the rail voltage; for a 42V rail 100-120V is fine for Vceo.
The higher the beta, the more the open loop gain of the amp, and the more is the feedback factor. Generally, this is good. So try to find a device no slower than 100MHz and a beta of 60 minimum.
I have used 2SC1819 and 2SC3955 with good results, but both these devices are now obsolete. Sanyo have a very good manufacturing process which gives high speed, low Cbc, and good beta. IMHO the voltage amp is the pivotal stage in a power amplifier. Watch the dissipation; no more than one third the rating for the device, and keep temperatures low for reliability.
Cheers,
Hugh
Thanks everybody!
I played a little bit with the KSA1381/KSC3503 (in the simulation) and they look fine... I think I will get them resp. the 2SA/2SC types. I had a look at the datasheet of the 2SA/2SC types and its pretty the same as for the KSA/KSC. So, I hope they will behave somewhat like the models.
With respect to the criteria given, Sanyo has some really nice types. But unfortunately I do not find any models for them. So, I will use what I have in my simulation and think of an update if the amp works as expected... BTW: Any chance of making a more or less good model out of the specs of a datasheet?
stegmaie
I played a little bit with the KSA1381/KSC3503 (in the simulation) and they look fine... I think I will get them resp. the 2SA/2SC types. I had a look at the datasheet of the 2SA/2SC types and its pretty the same as for the KSA/KSC. So, I hope they will behave somewhat like the models.
With respect to the criteria given, Sanyo has some really nice types. But unfortunately I do not find any models for them. So, I will use what I have in my simulation and think of an update if the amp works as expected... BTW: Any chance of making a more or less good model out of the specs of a datasheet?
stegmaie
Beware, Reichelt will often sell you second sources even if they list the real part number in the catalogue. As far as I know, the Japanese to not licence out the numbers, so any second sources are self appointed. I know of cases where there is a data sheet available for a Taiwanese second source, and about all that agrees is the wattage. I hate to think of those copies for which there aren't even published data sheets.
Best way is to order 1-2 pieces and if they are genuine, order more telling them what the code is.
Segor will do a very good job of telling you the manufacturer.
Best way is to order 1-2 pieces and if they are genuine, order more telling them what the code is.
Segor will do a very good job of telling you the manufacturer.
Thanks for the hint... I would have ordered from Reichelt...capslock said:
I don't know why, but I never liked Segor...
But, price is nearly the same (and in stock), so I will probably get them there. I have seen that the transistors have different ratings concerning Beta... I will ask them if its possible to get specific ones, resp. which ones they sell...
Thanks,
stegmaie
I bought a few sets of these a year or so ago when I purchase G. Randy Slone's book "the Audiophile Project Sourcebook" to build an amp in his book. He has used these in a lot of his projects but I can't seem to find them anymore, any suggestions.
Also does anyone know if there were any fakes made that I should watch out for if I find a supplier?
Thanks
Hi Upupa Epops,
Very fine devices, indeed! Looking at Sanyo BJT datasheets is pure semiconductor pr0n to me.
Do you know of any (EC) source, where you can order specific hfe classes? Or what do you usually get when ordering without hfe classification ocde?
Regards,
Peter Jacobi
Upupa Epops said:
Very fine devices, indeed! Looking at Sanyo BJT datasheets is pure semiconductor pr0n to me.
Do you know of any (EC) source, where you can order specific hfe classes? Or what do you usually get when ordering without hfe classification ocde?
Regards,
Peter Jacobi
Hi!
I asked Segor about manufacturers and ordering special classifications of the 2SA1381/2SC3503 and got the following reply (Sorry, only in German):
---
Vielen Dank fuer die Anfrage. Zunaechst ein wenigHintergrundinformation zur Marktstruktur, damit die Aussagenverstaendlich werden:Japanisch diskrete Halbleiter werden in Europa so gut wie gar nichtueber Liniendistributoren (=Vertrags-Grosshaendler derHalbleiterhersteller) vertrieben. Ausgenommen sind Grosstueckzahlenmit langen Lieferzeiten und einige wenige sehr haeufig eingesetzteTypen.Der weit ueberwiegende Teil dieser Bauteile wird von freienImporteuren aus Fernost bezogen - und dort auch nicht von denHerstellern direkt sondern wiederum ueber freie Haendler.Hierdurch treten zwei Schwachstellen zutage:- zum einen kommen immer wieder sog. Fakes vor. Das sind Faelschungen,die in der Stempelung den Originaltypen sehr aehnlich sehen, abernicht von den Originalherstellern stammen und daher oft die Datennicht vollstaendig einhalten. Die Faelschungsqualitaet reicht (wie beiBanknoten) von dummdreist (deutlich abgeschliffene Oberflaeche mitneuer fehlerhafter Stempelung) bis zu originalgetreu (optisch nicht zuunterscheiden).- zum zweiten gibt es eine Reihe von Sekundaerherstellern (SMT, WS,IEC, CDIL, ...) die die Originaltypen in mehr oder weniger guterQualitaet nachbauen. Im Unterschied zu Fakes werden sie aber alsSecond-Source-Ware verkauft und sind auch entsprechend gestempelt.Auch hier werden die technischen Daten des Originaltyps nicht immereingehalten, Datenblaetter der Secound-Source-Hersteller gibt espraktisch nicht.Unsere Politik zu diesem Thema sieht folgendermassen aus:- erkannte Fakes (wir haben einen Spezialisten, der das ziemlich guterkennt) werden dem Lieferanten zurueckgesandt oder verschrottet, derLieferant wird fuer diesen Artikel gesperrt.- wenn ein Japan-Halbleiter im Original zu bekommen ist, kaufenwir das Original. Nur bei Typen, die es nur noch von Second-Sourcesgibt, nehmen wir diese ins Programm.- wir kennzeichnen nach bestem Wissen und Gewissen in unserenArtikelbeschreibungen die Hersteller unserer Lagerware. Fehler sindaufgrund der Vielzahl der Typen nicht immer voellig auszuschliessen,wir wenden aber ziemlich viel Zeit fuer diese "Pflege" auf.Bei den von Ihnen angefragten Typen haben wir derzeit nach unsererKenntnis Sanyo-Original em Lager (mit der Restunsicherheit moeglicherFakes).Unsere generelle Empfehlung: fuer Entwicklungszwecke oderSerienfertigung sollten Sie nach Moeglichkeit Typen Europaeischer oderAmerikanischer Hersteller (Fairchild, Motorola/Onsemi, Philips, ...)einsetzen. Hierbei ist (bis zur Abkuendigung durch den Hersteller)eine kontinuierliche Lieferbarkeit von Originalware halbwegsgewaehrleistet. Bei den Japan-Halbleitern ist das nur bei wenigenHerstellern (z.B. Hitachi, Nec, Toshiba) fuer Grossmengenabnahmen mitlangen Lieferzeiten sicherzustellen.Aus den genannten Gruenden ist eine Bestellung von Selektionsstufenfuer Japan-Typen nicht moeglich - wir muessen nehmen, was derImporteur gerade am Lager hat (das sehen wir auch erst beimWareneingang).Mit freundlichen GruessenSEGOR-electronics GmbH- Dipl.-Ing. Th. Thiele -
---
stegmaie
pjacobi said:
I asked Segor about manufacturers and ordering special classifications of the 2SA1381/2SC3503 and got the following reply (Sorry, only in German):
---
Vielen Dank fuer die Anfrage. Zunaechst ein wenigHintergrundinformation zur Marktstruktur, damit die Aussagenverstaendlich werden:Japanisch diskrete Halbleiter werden in Europa so gut wie gar nichtueber Liniendistributoren (=Vertrags-Grosshaendler derHalbleiterhersteller) vertrieben. Ausgenommen sind Grosstueckzahlenmit langen Lieferzeiten und einige wenige sehr haeufig eingesetzteTypen.Der weit ueberwiegende Teil dieser Bauteile wird von freienImporteuren aus Fernost bezogen - und dort auch nicht von denHerstellern direkt sondern wiederum ueber freie Haendler.Hierdurch treten zwei Schwachstellen zutage:- zum einen kommen immer wieder sog. Fakes vor. Das sind Faelschungen,die in der Stempelung den Originaltypen sehr aehnlich sehen, abernicht von den Originalherstellern stammen und daher oft die Datennicht vollstaendig einhalten. Die Faelschungsqualitaet reicht (wie beiBanknoten) von dummdreist (deutlich abgeschliffene Oberflaeche mitneuer fehlerhafter Stempelung) bis zu originalgetreu (optisch nicht zuunterscheiden).- zum zweiten gibt es eine Reihe von Sekundaerherstellern (SMT, WS,IEC, CDIL, ...) die die Originaltypen in mehr oder weniger guterQualitaet nachbauen. Im Unterschied zu Fakes werden sie aber alsSecond-Source-Ware verkauft und sind auch entsprechend gestempelt.Auch hier werden die technischen Daten des Originaltyps nicht immereingehalten, Datenblaetter der Secound-Source-Hersteller gibt espraktisch nicht.Unsere Politik zu diesem Thema sieht folgendermassen aus:- erkannte Fakes (wir haben einen Spezialisten, der das ziemlich guterkennt) werden dem Lieferanten zurueckgesandt oder verschrottet, derLieferant wird fuer diesen Artikel gesperrt.- wenn ein Japan-Halbleiter im Original zu bekommen ist, kaufenwir das Original. Nur bei Typen, die es nur noch von Second-Sourcesgibt, nehmen wir diese ins Programm.- wir kennzeichnen nach bestem Wissen und Gewissen in unserenArtikelbeschreibungen die Hersteller unserer Lagerware. Fehler sindaufgrund der Vielzahl der Typen nicht immer voellig auszuschliessen,wir wenden aber ziemlich viel Zeit fuer diese "Pflege" auf.Bei den von Ihnen angefragten Typen haben wir derzeit nach unsererKenntnis Sanyo-Original em Lager (mit der Restunsicherheit moeglicherFakes).Unsere generelle Empfehlung: fuer Entwicklungszwecke oderSerienfertigung sollten Sie nach Moeglichkeit Typen Europaeischer oderAmerikanischer Hersteller (Fairchild, Motorola/Onsemi, Philips, ...)einsetzen. Hierbei ist (bis zur Abkuendigung durch den Hersteller)eine kontinuierliche Lieferbarkeit von Originalware halbwegsgewaehrleistet. Bei den Japan-Halbleitern ist das nur bei wenigenHerstellern (z.B. Hitachi, Nec, Toshiba) fuer Grossmengenabnahmen mitlangen Lieferzeiten sicherzustellen.Aus den genannten Gruenden ist eine Bestellung von Selektionsstufenfuer Japan-Typen nicht moeglich - wir muessen nehmen, was derImporteur gerade am Lager hat (das sehen wir auch erst beimWareneingang).Mit freundlichen GruessenSEGOR-electronics GmbH- Dipl.-Ing. Th. Thiele -
---
stegmaie
Stegmaie,
Thanks for posting that letter. Assuming my german was good
enough to get the essence of it, it seems there are reasons to
worry about buying japanese transistors. I get the impression
that what is said in that letter is likely to apply to all european
suppliers, not only the german ones. Not only are many transistors
hard to find, but those who do list them seldom or never tell
what manufacturer it is. According to this letter there seems to
be a great risk of getting questionable second-source products,
or even counterfeits (a well known risk already, I am afraid). I
have myself got CDIL-manufactured versions of some Hitachi
devices, and I still haven't even been able to confirm whether they
are even genuine CDIL (as an "answer" to somebodys question
earlier in this thread, these were 2SB649 or 2SD669, can't
remember which of them right now).
Is it worth the trouble trying to find suppliers for japanese
transistors, paying high prices for them and then risk getting
the wrong thing in the end anyway?
Many types can only be bought from smaller
companies, often selling them as spare parts, and these most
cetainly get their components through this long chain of distributor
mentioned in the letter, thus having little control over what they
actually get. Among those notoriously difficult-to-get types are
some of the popular Toshiba JFETS like J74 and K170. At least the
J74 is very difficult to get and I have found only a few suppliers
in Europe. What do you people get when buying these devices?
Are there second-sources out there? Are there counterfeits out
there?
Thanks for posting that letter. Assuming my german was good
enough to get the essence of it, it seems there are reasons to
worry about buying japanese transistors. I get the impression
that what is said in that letter is likely to apply to all european
suppliers, not only the german ones. Not only are many transistors
hard to find, but those who do list them seldom or never tell
what manufacturer it is. According to this letter there seems to
be a great risk of getting questionable second-source products,
or even counterfeits (a well known risk already, I am afraid). I
have myself got CDIL-manufactured versions of some Hitachi
devices, and I still haven't even been able to confirm whether they
are even genuine CDIL (as an "answer" to somebodys question
earlier in this thread, these were 2SB649 or 2SD669, can't
remember which of them right now).
Is it worth the trouble trying to find suppliers for japanese
transistors, paying high prices for them and then risk getting
the wrong thing in the end anyway?
Many types can only be bought from smaller
companies, often selling them as spare parts, and these most
cetainly get their components through this long chain of distributor
mentioned in the letter, thus having little control over what they
actually get. Among those notoriously difficult-to-get types are
some of the popular Toshiba JFETS like J74 and K170. At least the
J74 is very difficult to get and I have found only a few suppliers
in Europe. What do you people get when buying these devices?
Are there second-sources out there? Are there counterfeits out
there?
Christer:
I've been pretty lucky, so far
The small-signal (in TO-92 package) and medium-signal (like VAS drivers in TO-220 or TO-126 package) bipolars and small-signal jfets (also in TO-92 package) I've purchased over the years appear to have been legitimate products from the specified manufacturer (panasonic/matshushita, toshiba, hitachi, motorola --> on semi, nec).
I haven't heard or seen any complaints about counterfit small signal devices (yet)? Is this a problem also now?
The high power output devices I have and use were either:
1. Hitachi power MOSFETs purchased directly from Hitachi a long time ago - so no concern about fakes
2. Matshushita/Panasonic, Toshiba and Motorola/On-Semi bipolar and power MOSFET devices purchased long enough ago that fakes were not a concern yet
3. Samples directly from manufacturer - wouldn't be in their best interest to send me fakes ( or so I like to think )
As friend has shown me some fake Toshibas (3281, 1302) parts he purchased but I don't know where they came from.
Oh, I forgot, I bought a bunch of toshiba jfets from Borbely years ago and they definately were not fakes!
Frankly, I'm scared to buy any high power parts now unless they come straight from the manufacturer.
mlloyd1
I've been pretty lucky, so far
The small-signal (in TO-92 package) and medium-signal (like VAS drivers in TO-220 or TO-126 package) bipolars and small-signal jfets (also in TO-92 package) I've purchased over the years appear to have been legitimate products from the specified manufacturer (panasonic/matshushita, toshiba, hitachi, motorola --> on semi, nec).
I haven't heard or seen any complaints about counterfit small signal devices (yet)? Is this a problem also now?
The high power output devices I have and use were either:
1. Hitachi power MOSFETs purchased directly from Hitachi a long time ago - so no concern about fakes
2. Matshushita/Panasonic, Toshiba and Motorola/On-Semi bipolar and power MOSFET devices purchased long enough ago that fakes were not a concern yet
3. Samples directly from manufacturer - wouldn't be in their best interest to send me fakes ( or so I like to think
)As friend has shown me some fake Toshibas (3281, 1302) parts he purchased but I don't know where they came from.
Oh, I forgot, I bought a bunch of toshiba jfets from Borbely years ago and they definately were not fakes!
Frankly, I'm scared to buy any high power parts now unless they come straight from the manufacturer.
mlloyd1
Mlloyd1,
Yes, I agree I haven't so far heard about any counterfeit small signal
devices either. I would not jump to the conclusion that there aren't
any, though, but yes, probably they are rather safe to buy. Maybe
there is too little money in faking those.
I also suppose there may be some difference between US and Europe.
US is one very large market, so you have plenty of distributors large
enough to buy directly from the manufacturers, I suppose, while
many European distributors are smaller and play on single-country
markets mainly, making it more difficult for them to buy directly from
manufacturers, except for devices sold on a very large scale. Many
of the japanese types are difficult or impossible to get from the
bigger supplisers, so I think a cautious attitude is sensible. For
instance, our major swedish distributor Elfa (selling also in the rest
of Scandinavia and the around the Baltic) may be large enough to
buy from manufacturers, and do have agreements with many
manufacturers. However, they have been sadly cttring down on
many transistor types in recent years in very strange ways. They
continue to sell 2SK170 but has stopped supplying its P-channel
complementary 2SJ74. They sold the complementary drivers
2SA1837 and 2SC4793(?), but for some puzzling reason
now sell only the PNP one. They sold both Philips and ON versions
of BD139/140 but have now stopped selling the Philips version
which were the only ones having at least the most essential
datasheet info.
Now that the dollar is to our advantage, I might consider ordering
from the US despite the additionl customs fees (not always added).
However, I have no idea which of the US distributors, if any, are
reliable in this respect. I actually checked both Mouser and Digikey
but couldn't even find the Toshiba JFETs mentioned above.
Yes, I agree I haven't so far heard about any counterfeit small signal
devices either. I would not jump to the conclusion that there aren't
any, though, but yes, probably they are rather safe to buy. Maybe
there is too little money in faking those.
I also suppose there may be some difference between US and Europe.
US is one very large market, so you have plenty of distributors large
enough to buy directly from the manufacturers, I suppose, while
many European distributors are smaller and play on single-country
markets mainly, making it more difficult for them to buy directly from
manufacturers, except for devices sold on a very large scale. Many
of the japanese types are difficult or impossible to get from the
bigger supplisers, so I think a cautious attitude is sensible. For
instance, our major swedish distributor Elfa (selling also in the rest
of Scandinavia and the around the Baltic) may be large enough to
buy from manufacturers, and do have agreements with many
manufacturers. However, they have been sadly cttring down on
many transistor types in recent years in very strange ways. They
continue to sell 2SK170 but has stopped supplying its P-channel
complementary 2SJ74. They sold the complementary drivers
2SA1837 and 2SC4793(?), but for some puzzling reason
now sell only the PNP one. They sold both Philips and ON versions
of BD139/140 but have now stopped selling the Philips version
which were the only ones having at least the most essential
datasheet info.
Now that the dollar is to our advantage, I might consider ordering
from the US despite the additionl customs fees (not always added).
However, I have no idea which of the US distributors, if any, are
reliable in this respect. I actually checked both Mouser and Digikey
but couldn't even find the Toshiba JFETs mentioned above.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.